By
(I've been saying that January is the sweet spot, but as many of you know, I actually believe that late December is the sweetest spot. We don't have daily or weekly data so there is no way to actually test my theory, but the Foolish Four Portfolio renews during the last week of December. That lets us avoid the volatility that often hits on the first trading day of the year. Also, since we set the portfolio up to work as a taxable account, it renews after a year and a day, so it wouldn't be possible to renew on the first trading day of the year, ever year, anyway. The solution was to start in December and renew one day later each year until we either slip into January or have a bad year, at which time we can renew early since the tax difference would be minimal.)
Fool on and prosper!
According to the pundits, the market is simply suffering from interest rate jitters because the economy is growing so strongly. That doesn't strike me as serious unless you are a day trader who forgot to go short this morning, so let's move on to something interesting.
Bob Price (TMF Sandy) has just completed a study of the Foolish Four strategy using a two-year holding period instead of a one-year hold, and the results are quite striking and rather surprising.
When we looked at our calendar year spreadsheet (January returns only) and changed the holding period to two years instead of one (every stock is held for two years before even thinking about renewing it), we found that the average annual return (CAGR) dropped a bit, but so did the standard deviation.
Perhaps a bit more on standard deviation is in order. Remember that a long-term average of 20% can include years where the return was 40% or more and years where the return was -20% or less. Standard deviation measures how much variation there is in the numbers that make up the "average." So a lower standard deviation means that the actual yearly returns were closer to the average annual return. (For more on standard deviation see "Risk and Reward.")
So an ideal strategy will have high returns and a low standard deviation. That's not usually the case, but it's something to look for, and when comparing strategies, a lower standard deviation is definitely a plus.
When we found that the returns for two-year portfolios started in January were lower (bad) but the standard deviation was also lower (good), that led us to assume that a two-year hold might be a good choice for someone who wants a more stable return and is willing to trade a few percentage points of total return to get it. Who would want that? Retirees and the risk-adverse both come to mind, as well as those who have a truly Foolish inclination toward long-term buy-and-hold strategies.
Now, thanks to Bob Price's study we have a much more detailed look at two-year strategies, and, surprise! things are not what we thought. The January data didn't change -- the return still drops if you hold your stocks for two years instead of renewing them after 12 months, and the standard deviation also drops. So far, so good. But when we look at portfolios starting in other months, the story is quite different.
For 10 of the 11 other months, returns were higher when you held all of your stocks for two years, and for the first half of the year, so was the standard deviation. This is exactly the opposite of what we found for the January portfolio. So much for nice, neat hypotheses.
What can we make of this? Well, the first thing that struck me was that the advice to "hold until the end of the next year if you started in a month other than December or January" was better than I thought. (Sometimes you just get lucky.) I've been suggesting that people who start in the middle of the year hold through their first year anniversary and plan to rebalance in December or January of the following year because I knew that short holding periods tended to be very volatile and because this way they would get into the sweet spot for future portfolio renewals. The two-year study certainly backs that up.
But even though the returns are higher for two-year portfolios except for those started in January, they are not high enough to suggest simply sticking with the month you started in. For example, the average return for portfolios renewed in July was 14.35% with a one-year hold and 15.12% with a two-year hold. The two-year hold is better but still far below January's return of 19.45%. (These are all 38-year averages, 1962-1999. The 25-year averages are several percentage points higher since they don't include the uncertain '60s and the recession of 1973-74.)
Once again January is the sweet spot. And since the returns for two-year, January portfolios are considerably lower (16.06% vs. 19.45%), the standard one-year, January portfolio still looks like the best bet. Even though the standard deviation is higher for the one-year hold, it's not higher enough to offset the difference in the returns. For other months, the standard deviations are not significantly different. The idea that a two-year hold would provide more stable returns appears to be a myth.
Here are the returns so you can see for yourself.
1 year 2 year
Month CAGR CAGR Difference
Jan. 19.45% 16.06% -3.39%
Feb. 16.10% 16.33% 0.23%
Mar. 15.97% 16.59% 0.62%
Apr. 14.49% 15.37% 0.87%
May 14.05% 14.37% 0.32%
June 13.02% 13.83% 0.81%
July 14.35% 15.12% 0.78%
Aug. 12.71% 14.56% 1.85%
Sept. 15.46% 15.10% -0.36%
Oct. 15.30% 16.39% 1.09%
Nov. 17.14% 17.66% 0.52%
Dec. 17.89% 18.36% 0.47%
Average 0.66%
1 year 2 year
Month STDEV STDEV Difference
Jan. 0.19360 0.17302 -0.02058
Feb. 0.16939 0.19239 0.02300
Mar. 0.18308 0.18471 0.00162
Apr. 0.17513 0.20414 0.02902
May 0.19640 0.20327 0.00687
June 0.18287 0.19192 0.00905
July 0.18645 0.19738 0.01093
Aug. 0.23401 0.21805 -0.01596
Sept. 0.23130 0.23022 -0.00107
Oct. 0.24458 0.23267 -0.01191
Nov. 0.24663 0.22422 -0.02240
Dec. 0.26939 0.26566 -0.00373
Average 0.00040
On Monday we will take a closer look at the relationship between the returns and the standard deviation. (Just what you were hoping for!)